Citalopram exposure of hESCs during neuronal differentiation identifies dysregulated genes involved in neurodevelopment and depression

Abstract

Selective serotonin reuptake inhibitors (SSRIs), including citalopram, are widely used antidepressants during pregnancy. However, the effects of prenatal exposure to citalopram on neurodevelopment remain poorly understood. We aimed to investigate the impact of citalopram exposure on early neuronal differentiation of human embryonic stem cells using a multi-omics approach. Citalopram induced time- and dose-dependent effects on gene expression and DNA methylation of genes involved in neurodevelopmental processes or linked to depression, such as BDNF, GDF11, CCL2, STC1, DDIT4 and GAD2. Single-cell RNA-sequencing analysis revealed distinct clusters of stem cells, neuronal progenitors and neuroblasts, where exposure to citalopram subtly influenced progenitor subtypes. Pseudotemporal analysis showed enhanced neuronal differentiation. Our findings suggest that citalopram exposure during early neuronal differentiation influences gene expression patterns associated with neurodevelopment and depression, providing insights into its potential neurodevelopmental impact and highlighting the importance of further research to understand the long-term consequences of prenatal SSRI exposure.

Keywords: DNA methylation; citalopram; depression; epigenetics; human embryonic stem cells; multi-omics; neurodevelopment; single-cell RNA-seq.

FIGURE 1

Neurodevelopmental effects of citalopram in a model of early human neuronal development. (A) Schematic representation of neuronal differentiation of hESCs. The cells were continuously exposed to media only (control cells) or 50, 100, 200 or 400 nM citalopram from Day 1 and throughout the differentiation process. Samples were collected for multi-omics analyses at Day 0, 6, 10 and 13. (B) Viability of hESCs after citalopram exposure for 24 h. (C) The presence of the stem cell marker OCT4 and neuronal markers SOX1, TUBB3 and NCAM1 was assessed by flow cytometry at Day 0 and Day 13 in control cells and cells exposed to 400 nM citalopram. ΔMFI, delta median fluorescence intensity.

FIGURE 2

Time-response of citalopram-exposure during neuronal differentiation of hESCs. (A) Gene expression (GE) levels (log counts per million, logCPM) of selected serotonin receptor genes for Day 6 to Day 13. (B) Table showing the number of genes and CpGs that responded to the different concentrations of citalopram over time (Day 6 to Day 13). Venn diagram showing overlapping DEGs and DMGs. (C) GE levels (logCPM) of selected linear time-response DEGs. (D–F) Shared biological processes (BPs) among genes that were differentially expressed in cells exposed to (D) 100 nM, (E) 200 nM or (F) 400 nM citalopram over time. (G) GE levels (logCPM) of selected non-linear time-response DEGs. Genes with adjusted p-value <0.05 were considered significant.

FIGURE 3

Overlapping differentially expressed and methylated time-response genes. Gene expression levels (logCPM) and DNAm levels (beta values) of selected overlapping DEGs and corresponding DMCs.

FIGURE 4

Citalopram-exposure affects gene expression in a dose-dependent manner during neuronal differentiation. (A) Number of DEGs and DMCs identified in the dose-response analysis. Venn diagram showing overlapping DEGs and DMGs. (B) Top shared BPs between genes that responded to citalopram in a dose-dependent manner at Day 13. (C) GE levels (logCPM) of selected linear dose-response DEGs. (D) GE levels (logCPM) of selected non-linear dose-response DEGs. Genes with adjusted p-value <0.05 were considered significant.

FIGURE 5

scRNAseq reveal small differences in cell type composition between cells exposed to citalopram and control cells. (A–D) All cells projected in UMAP plots colored by (A) Day, (B) exposure group, (C) slingshot pseudotime and (D) annotation to LaManno brain data cell types. (E) Proportion of cells in each exposure group annotated to cell types according to the LaManno brain and Bhaduri organoid datasets. eSCa-c; hESCs, eProg1-2; hESC-derived neuronal progenitor cells, IPC; intermediate progenitor cells. (F) UMAP plot of all cells colored by Seurat clusters and (G) corresponding proportion of cell of each exposure group annotated to clusters. (H) Single cell gene expression of top ten overlapping DEGs between cells exposed to different concentrations of citalopram compared to control cells at Day 6, Day 10 and Day 13.

FIGURE 6

Temporally expressed genes respond differently in citalopram-exposed cells compared to control cells during differentiation. (A) Pairwise comparisons between citalopram and control for RNA-seq (DEGs), scRNAseq (scDEGs) and DNAm (DMCs) at Day 6–13. (B) Boxplot of Slingshot pseudotime values per exposure group for Day 6–13. Significant comparisons are marked with asterisks (Student’s t-test, ****: p ≤ 0.0001). (C) GE levels (logcounts) of DDIT4, MTHFD2, HES5, FRZB, LRATD2, NNAT, EFNB2, TAGLN, NR2F1 and SESN2 selected from the top 100 temporally expressed genes, which responded differently in citalopram-exposed and control cells across Slingshot pseudotime at Day 6–13. Each dot represents one cell, and the lines represent the average GE for each citalopram concentration across pseudotime.